Durba Sengupta

Science Enthusiast and Researcher

[#WOMENSART @womensart1]

Persepolis (2000), autobiographical graphic novel by Iranian cartoonist, illustrator, author and film director Marjane Satrapi #WomensArt ❤️ RIP

[Ingo Hartung @HartungIngo]

It‘s time to remove the question mark!

A huge congrats to @CraigMCrews, Ray Deshaies and the scientists @ArvinasInc.

1st FDA approval for a #PROTAC on May 1st 2026. A day to remember. Extra drinks at the Proxidrugs GRC in a few weeks!

[Alisia Fadini @FadiniAli]

Statistical crystallography as a route to understanding protein dynamics! 1,146 crystal structures of SARS-CoV-2 Mpro reveal correlated motions connecting the dimer interface to the active site. Great to see this work out by the DESY team.
Link: https://t.co/6lClrhfzW8

[Jorge Bravo Abad @bravo_abad]

Flow matching is emerging as a unifying framework for generative biology

Biology is full of mappings between states: a healthy cell turning diseased, amino acids folding into a functional protein, a ligand docking into its target. Deriving such transformations analytically is intractable—which is where generative AI steps in, and flow matching is quickly becoming its backbone.

Morehead and coauthors review how flow matching (FM) is reshaping generative modeling in bioinformatics. Unlike diffusion models, FM doesn't force the source distribution to be Gaussian: it learns a time-dependent vector field that transports samples between any two distributions along straight-line, optimal-transport paths. The payoff: fewer inference steps, simulation-free training, and built-in support for geometric priors like SE(3) equivariance—essential for 3D biomolecules.

What's striking is how fast FM has spread across biological scales. For molecules, FoldFlow, FrameFlow, and Multiflow generate protein backbones on SE(3)ᴺ manifolds, SemlaFlow produces valid small molecules up to 100× faster than diffusion, and Dirichlet FM handles discrete DNA/RNA sequences. FlowDock and NeuralPLexer3 predict protein–ligand complexes that match or exceed AlphaFold 3 on key benchmarks, while AlphaFlow and MDGen generate conformational ensembles and MD trajectories. At the cellular scale, CellFlow and Meta FM map unperturbed populations to perturbed states, and CryoFM and FlowSDF extend FM to cryo-EM and microscopy.

The deeper point: FM subsumes diffusion models, continuous normalizing flows, and optimal transport as special cases, providing scaffolding for an AI-based virtual cell—simulating molecular, structural, and phenotypic effects of perturbations across scales.

Overall, this signals a shift in what's computationally tractable. Instead of narrow, stage-specific models, FM points to unified conditional generators that design sequences, predict complexes, and model perturbation responses in one framework—shortening wet-lab cycles and making closed-loop, active-learning workflows practical.

Paper: Morehead and coauthors, Nature Machine Intelligence (2026) — Journal license | https://t.co/7UyfTWXKmS

[João Pedro de Magalhães @jpsenescence]

Who did this? 😂

[Historical Africa @historical_Afr]

Respect!!

[Durba Sengupta @durbignon]

@arshukla @IITKanpur Congratulations to you and your amazing team!

[Natalie Khalil @natalienkhalil]

Basically

[manalisjoshi @manalisjoshi1]

Excited to share our work on β1AR activation! Unique microswitches in extracellular regions, trigger the common GPCR activation mechanism. Led by @shrutikoulgi, with Uddhavesh Sonawane, @durbignon and in collaboration with Daniel Nietlispach and team.
https://t.co/KGtqy0qJqY

[Sumana Roy @SumanaSiliguri]

My passion project is out today, in the US, UK, Europe, Australia ...

About the book from the Yale University Press website:

https://t.co/y0TSZBlBFG

[Sandhya Koushika @WormlockHolmes]

🙏🏾🙏🏾🙏🏾🙏🏾 to the hard work & fantastic organization for #OBMT26 to
Front row L to R: @Ssaleem_M Abdur Rahaman @kuruhoornambi

Back row L to R trainees: Urbi Ghosh, Divit Solanki, Aninda Sundar, Debraj Koiri, Gargi Dey, Riddhisha Patra, Neelrathan, Srijita Roy, Swapnil Sahoo

[Sandhya Koushika @WormlockHolmes]

Coarse grain Martini molecular model of Caveolin-binding shows that Caveolin both induces and senses membrane curvature. Caveolin binding induced curvature shows that cholesterol & PS are found in the neck. @durbignon @csir_ncl #OBMT26

[Durba Sengupta @durbignon]

Becoming more critical with all the cylinder shortages

[Raghunath Mashelkar @rameshmashelkar]

National emergency requires emergent actions.

When global chokepoints threaten LPG supplies, Indian innovation has the power to secure sovereignty.

India’s own @CSIR-NCL Dimethylether (DME) technology can power kitchens with indigenous fuel. 🇮🇳

DME is a substitute for LPG with the advantage of cleaner combustion, lower NOx & Sox and  greater strategic resilience.
It can be produced from methanol derived from India’s own coal or biomass.

Indian science, Indian fuel, Indian kitchens: energy atmanirbharta.

.
@csir_ncl has successfully demonstrated the indigenous DME technology at 250 kg/day and partners are ready to build a 2.5 TPD demonstration plant quickly.

 It needs support from Centre for High Technology (CHT)  https://t.co/f2GplOAdbV, which has the funds precisely for such projects.

The Government is fully seized with the emergent situation.

https://t.co/TUBZgIjvG1

As a former Chairman of SAC to Ministry of Petroleum and Natural Gas, I have myself seen how CHT has helped accelerate development of indigenous technology.

I urge  CHT to take emergent steps to immediately support scale up to deal with this national emergency.

@PMOIndia @PetroleumMin @DrJitendraSingh @PrinSciAdvOff

[Oxford Mathematics @OxUniMaths]

So who discovered the Fibonacci sequence? Fibonacci?
No, afraid not (no disrespect Leonardo Bonacci, aka Fibonacci).
Another scientist?
Nope.
Who then?
Here's @MarcusduSautoy (it wasn't him either in case you're wondering).

[Prof. Nikolai Slavov @slavov_n]

We have known that proteins are made of amino-acids for ~ 120 years.

In 1902, Fischer and Hofmeister proposed that the -NH2 of one amino acid binds the -C00H of another, resulting in a linear structure that Fischer termed "peptide".

[Chattopadhyay Lab @amit_gpcr]

Happy to share our recent paper https://t.co/RtoLg5tWe6 where we computationally analyzed the specificity of the serotonin1A receptor-PIP lipid interactions using multi-microsecond long coarse-grain MD simulations.@durbignon @ccmb_csir @CSIR_IND @ACSChemNeurosci #gpcr

[Durba Sengupta @durbignon]

Yet another new index! But this time calculating "mentorship"...

My M10-index is 36!

Find yours at: https://t.co/JnVzYDLzXO

[Ragothaman Yennamalli @StructBioinfo]

Interesting!!!!

The Mentorship Index (M-index) measures a scientist's contribution to mentoring junior scientists. My M10-index is 21! Find your Mentorship Index: https://t.co/hXrannS09R

[Durba Sengupta @durbignon]

A real issue in science administration here...

[Leah Pierson @leah_pierson]

omg this title, this paper

[World of Statistics @stats_feed]

Dmitri Mendeleev, who is credited with creating the Periodic Table of Elements was nominated for a Nobel Prize in Chemistry 9 times but never won.

His awarding was blocked each time by the 1903 winner, Svante Arrhenius, who held a grudge against Mendeleev for criticizing one of his papers.

[Anand Srivastava @mbuanand]

We are organizing a 7-day EMBO sponsored international workshop at @iiscbangalore @MBU_IISc (April 26-May 03, 2026). "AI-driven approaches in Structural Biology". Link: https://t.co/l1tsv19sbH.

Please retweet. Looking forward to seeing some of you in Bangalore!